Voltage-controlled amp design is hard

And the rest of us have some idea who that might be. ;)

The initial application is for retrofitting existing instruments, which use a voltage-output DAC to control the PMT gains via a control input on the PMT module. Rewriting their software to suit a fixed-gain TIA, or to use a PGA, would make life much more complicated.

Once they've shipped some, and are comfortable turning off their existing vendor, subsequent designs will probably just use a PGA like normal people. ;)

Meanwhile in order to get my gizmo into their instruments, I need to make them something that works like their purchased module.

Cheers

Phil Hobbs

Darn you. Beat me to it.

uP with internal ADC, driving a serial MDAC, might work.

You could cheat and have one or so switchable gain stages, spliced into the MDAC function, to improve s/n over the dynamic range. Save a couple of cal factors in the uP flash.

Yup, that's what I'm proposing. Or possibly a dual dpot--the gain of a PMT goes as some high power (like the 8th) of the supply voltage, which would b e easier with both the input and fb resistors variable. 300 kHz is about wh ere capacitive feedthrough in MDACs becomes important at low codes. I reall y don't want pulse-shape artifacts.

Not a bad idea. Since I'm not building the test stand, the cal constants wo uld be hard-coded. The gain doesn't have to be that accurate, fortunately, but it would be good for it to be settable at the 1% level. With such a no nlinear curve, it's easy to run out of bits if you're not careful.

Cheers

Phil Hobbs Phil Hobbs